Perspectives on Human Memory and Cognitive Aging

Essays in Honor of Fergus Craik

Moshe Naveh-Benjamin, Morris Moscovitch, Henry L. Roediger, III

• Publisher: Psychology Press
• Year: 2013
• eBook ISBN: 9781134949694
• Edition: 1
• Topic: Psychologie
• Subtopic: Kognitive Psychologie & Kognition

Divided into four parts, the first section of this book deals with levels of processing and memory theory, the second addresses working memory and attention, the third deals with cognitive aging, and the last addresses neuroscience perspectives.

Part

I

Levels of Processing and Memory Theory

1

Chapter

Michael J. Watkins

Part I Introduction: Levels of Processing

A little more than three decades ago, a singular young researcher joined his students at lunch and mentioned that he had been thinking. His students harkened at once, for Fergus Craik is a modest man, not given to bluster. Craik reached for a scrap of paper and drew a horizontal line, from which dropped two vertical lines, one short and one long.

“What’s that, Gus?”

“It seems to me,” said he, “that memory for something depends on how deeply the something is processed or analyzed.”

“That’s it, Gus?”

“Well, that’s the essence of it…”

The notion was at once bemusing and appealing. The drawing seemed too insubstantial and too wobbly to compete with the solid boxes-and-arrows drawings that were the hallmark of the dazzling new information processing era. Yet, it depicted an idea that was both simple and plausible.

The mixed reactions of Craik’s students foreshadowed those of the wider world, as is aptly attested in the five chapters that follow. In the first two of these chapters, the levels-of-processing proposal is subjected to such weighty arguments and so many recalcitrant findings that the reader might be forgiven for assuming the proposal to be now perfectly dead. But even in a field not known for swift and certain demise, this proposal has an unusual penchant for resurrection, and sure enough, with the next three chapters it is back in fine fettle, shaping the conduct of research and the digestion of its fruits.

Why do some argue so passionately for fundamentally revising the levels-of-processing proposal or even consigning it to history while others continue to embrace it? Part of the answer may lie in a general disinclination to see the proposal as itself submitting to different levels of analysis. From the start, the authors gave it a metatheoretical status on the premise that it is not a theory but a framework (Craik & Lockhart, 1972; see also Robert Lockhart’s discussion in chapter 7), but not everyone has accepted the validity of this premise. Perhaps it should be regarded as half valid. Thus, the notion that memory is a byproduct of perceptual-cognitive analyses constitutes a general perspective on memory or an approach to its study, and framework serves well enough for this purpose. But the notion that these analyses have an inherent depth that determines memory durability is a theory by almost any definition.

The first two chapters of this tribute address a fundamental limitation of the first, or metatheoretical, component: If an event of time 1 is recalled at time 2, then the levels-of-processing proposal addresses only time 1 and so cannot depict the complete memory process. In particular, it is silent on any effect involving type of test. As Lockhart notes in chapter 7, Craik and Lockhart (1972) willingly conceded this limitation from the start. And yet, to this day, it troubles certain prominent memory theorists. In chapter 2, Endel Tulving exhorts Craik to extend his consideration at least a little way beyond time 1, and in chapter 3 Henry Roediger and David Gallo argue that consideration should be extended all the way to time 2. At play here is the abiding convention that times 1 and 2 frame a unitary mechanism, their temporal disparity being bridged by a memory “trace.” From this perspective, an account of memory confined to time 1 is no more complete than an account of the telephone system confined to the mouthpiece. The reader may do well to keep in mind, however, that, absent grander pretensions, an account of a telephone mouthpiece may, of itself, be entirely valid and potentially useful.

Tulving marshals evidence for dissociating the perceptual and cognitive encoding addressed by the levels-of-processing proposal from subsequent encoding that is assumed to occur in trace formation, and he affectionately castigates Craik for neglecting the latter. And it is no defense, Tulving adds, that some of the evidence he offers comes from beyond the realm of psychology, for cognitive psychology can no longer be pursued in isolation from the brain sciences. Here Tulving is unquestionably tapping into conventional wisdom and, having acceded to this wisdom, Craik may indeed appear vulnerable to the charge of leaving the trace underburdened in his conception of memory. One possible response to this difficulty would be to retreat to a radically functional position whereby recall and its manifestations in various memory tests at time 2 are a function of analyses that occurred at time 1, with the idea of an intervening memory trace inaccessible to psychological enquiry being rejected as the subject of natural science rather than of psychology. A more likely response is that the levels-of-processing proposal was never intended to explain all memory findings.

Roediger and Gallo focus on the showcase paradigm for levels-of-processing research, in which the effects on memory are compared for items subjected to two or more tasks designed to restrict processing to different depths. Specifically, they review several findings that collectively, if not individually, seriously undermine or even disconfirm a strict reading of the levels-of-processing proposal, both as metatheory and as theory. But taken literally, the proposal was a nonstarter. For example, that orienting tasks do not always constrain processing to the designated level is clearly demonstrated by the venerable Stroop effect, wherein the meaning of a word may affect the naming of the color in which it is written. Perhaps the moral of Roediger and Gallo’s review, then, is merely that the proposal should not be taken literally; certainly, this moral has been studiously respected by the authors. Nevertheless, the levels-of-processing proposal has surely increased the significance of noncompliant findings, and to this extent, at least, it has infiltrated our collective wisdom.

Such infiltration is transparent in the remaining three chapters of Part I. Boris Velichkovsky argues that the levels-of-processing concept may turn out to be just the psychological aspect of a grander dynamic system that also includes the activity and evolution of eye and brain. John Gardiner, Alan Richardson-Klavehn, Cristina Ramponi, and Barbara Brooks show that, contrary to conventional wisdom, an effect of processing level may extend to incidental, or involuntary, remembering. And Betty Ann Levy argues from research on the acquisition of reading fluency, and specifically on the conditions under which the reading of a text is facilitated by a specific prior reading, for the relation between processing and memory being, not one way, but reciprocal.

The chapters in Part I increment a Craik and Lockhart (1972) citation count that is, as Roediger and Gallo document, astonishing by any measure. And, as Lockhart predicts, its core proposal will continue to color our thinking—even as it draws sustenance from would-be obituaries—for many years to come.

Reference

Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11, 671–684.

2

Chapter

Endel Tulving

Does Memory Encoding Exist?

Does memory encoding exist? What a silly question—of course it does! The whole world’s literature on memory is full of papers and books on encoding, or phenomena involving encoding, and surely one cannot write about something that does not exist, at least not in science. Or can one?

At one time physicists wrote at great length about the aether, chemists about phlogiston, and biologists about élan vital, and everything that was said about these things turned out to be wrong, because they did not exist. While pursuing their interests, scientists did find out many things about their subject matter that became not only useful but true, but the fact remains that it is quite possible for scientists to think, talk, and write about nonexistent entities. Is encoding one of them?

There are two basic kinds of existences in the world: physical and mental. Many very clever people, from Descartes to the great thinkers of our own age, have written a great deal about the nature of the relation between the physical and the mental. Physical existences include things such as galaxies, stars, planets, oceans, pebbles on the seashore, tortoises, and worms, and the molecules and atoms that they are made of. Mental ones consist in emergent creations of an unbelievable physical thing, the brain: percepts, images, feelings, ideas, beliefs and the like, and their equivalents in nonhuman animals.

Whether or not encoding exists in the physical world is as yet unknown, but encoding certainly exists as a thought in the minds of many people. Encoding is a concept. In the science of memory, we have a very large number of concepts. We use them continually and are utterly dependent on their existence. We have linguistic, or sometimes mathematical, labels for them, terms as these labels are called, very much in the same way that we have symbolic labels for things that do exist as tangible physical entities. We create concepts and terms, describe and define and explain them, have faith in them, defend them against nonbelievers, and measure the progress in the understanding of our subject matter by evaluating the usefulness and validity of our terms and concepts.

Now, encoding may exist as a concept, but the bigger question is whether encoding also exists in the world, outside some individuals’ imaginations. Is it a real part of something that happens even if there are no sentient beings who think about it? A falling tree in the woods makes no sound if there is no living soul around to hear it—sound is the product of an appropriately endowed nervous system—but it does produce compressions of the air regardless of the presence of witnesses.

It is appropriate to raise the question about the existence of encoding in the present context because our friend and colleague, Fergus Craik, whose brilliant scientific career this volume is meant to celebrate, has repeatedly expressed doubts about such existence. Others may use the term designating the concept, but he does not believe it corresponds to anything in reality. Encoding is, our friend implies, memory science’s élan vital, a thought in the minds of some thinkers, but it does not exist in reality. What does exist is perception, different kinds of perception at that, and the term encoding is mistakenly applied to some of these kinds.

Is our friend right? If so, it follows that those others who believe in encoding must be wrong. The question of who is right and who wrong is obviously of some interest, because we cannot expect to get very far in our pursuits of the truth about memory unless we solve the encoding problem first. I am not suggesting that the encoding problem is the only hurdle to be crossed before progress in our understanding of memory can occur, but I am suggesting it is among the very basic ones.

The issue before us, then, is this: Is there something in nature that corresponds to the idea that (many) students of memory have and that they designate as encoding and that is different from perception? Tackling the issue also means asking how we can find out. How do we go about establishing the existence of something that is invisible, inaudible, and intangible, and how do we determine its “basic” difference from another such? And, most germane to the occasion, is there a way of persuading Gus Craik to accept encoding as a part of reality?

This essay consists of five parts:

1. A brief history of the science of memory, the concept of encoding, and Gus’s place in it.
2. Byproduct theory of trace formation: What is it and why is it postulated?
3. GAPS (general abstract processing system): Summary of the opposing view. What is it, how does it differ from the byproduct theory, and why is it postulated?
4. Functional neuroimaging and the novelty encoding hypothesis: Fleshing out the encoding’s skeleton.
5. These four main sections are followed by a concluding statement.

A Brief History of the Science of Memory

As this is a historic occasion, let us begin by placing our friend’s thoughts about encoding into some kind of a historical perspective. The following thumbnail sketch of the history of the psychological science of memory, borrowed from Tulving (1993), is extremely brief and undoubtedly biased, but it serves the current purpose. For a fuller and more objective view, see Bower (2000).

The short history of the study of human memory can be divided into four successive stages, each stage bringing in innovations not known in the preceding one. The stages are shown in Table 2.1. The first one began in antiquity and ended in 1885. In that year, the second stage was launched by the publication of Herman Ebbinghaus’s Über das Gedächtnis (Ebbinghaus, 1885). It lasted some 75 years, to around 1960. During that time the field came to be dominated by English-speaking scientists, especially in America. Problems of memory were pursued by experimental psychologists under the general rubric of verbal learning. Most of the research activity had to do with precise measurement of basic phenomena of learning and forgetting of lists of verbal materials under tightly controlled experimental conditions. The central theoretical concepts were association and its single property, strength. Theories consisted of attempts to explain observed facts in terms of acquisition, retention, and transfer of, interference with, and mediation by associations. In America, the dissenting voice of Frederic Bartlett (1932) from across the Atlantic was not heard.

Table 2.1. Four successive stages of the science of memory

Metaphysics	From the early Greeks on
Verbal learning	After 1885
Information processing	After circa 1960
Cognitive neuroscience	After circa 1980

In the years around 1960, the associative verbal-learning framework saw a rival emerging in the form of the information processing paradigm. A much wider variety of problems, issu...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright
• Contents
• Preface
• List of Contributors
• Fergus Craik: A Biographical Sketch
• Part I: Levels of Processing and Memory Theory
• Part II: Working Memory and Attention
• Part III: Age-Related Changes in Memory and Cognition
• Part IV: Neuroscience Perspectives on Memory and Aging
• Author Index
• Subject Index
• Figure 16.1